Optimization study of heating performance for an impinging jet ventilation system based on data-driven model coupled with TOPSIS method

Abstract

Impinging jet ventilation (IJV) is believed to create good indoor thermal comfort and air quality (IAQ) in an energy efficient way. However, since there is little discussion about the relationship between its heating performance and airflow patterns, operation control of ventilation parameters (e.g., supply temperature, airflow rate, discharge height and nozzle shape) is not clear. To achieve a good heating performance for IJV, operation conditions are optimized by using data-driven model coupled with technique of order preference by similarity to ideal solution (TOPSIS). Draught discomfort (PD), air change efficiency (ACE) and energy utilization coefficient (EUC) are modeled as performance indicators. First, response surface methodology (RSM) and artificial neural network (ANN) are employed to establish predictive models for the performance indicators, a comparison between the two methods are then performed from prediction error analysis. It shows that both RSM and ANN can predict the ACE and EUC well, but compared to ANN, RSM is less reliable in estimating PD and may fall into the risk of misjudging comfort level. Then, the ANN is used together with the TOPSIS to optimize the IJV for heating and a square nozzle is better preferred among the studied nozzle shapes. Moreover, optimal conditions of supply parameters are determined and verified for different nozzle characteristics. Overall, the current results not only can provide reference for the design and operation control of ventilation parameters of IJV for heating, but also contribute to the practical applications of ANN in different ventilation scenarios.